LONDON, 4 October 2004 — Today, as always, the unpredictability in currency trends poses challenges to asset managers in the selection of various currencies that will be represented in their portfolios.
Besides its fundamental importance, the choice of currencies is unavoidable.
It is true that currency choice cannot be avoided by portfolios which appear to have domestic character.
With the development of large-scale markets in all the major currencies, any asset manager charged with the task of maximizing return must regard non-domestic currencies as sources of investment opportunities which must, at least, be investigated in terms of risk and return.
Therefore, the choice of currencies cannot be avoided by asset managers, except those who, for reasons of exchange control, are not permitted to invest in currencies other than their own. The choice of currencies presents itself as an unavoidable problem of very great importance.
This would present little difficulty if there were generally accepted logical principles which could be used in its solution. Unfortunately this is not the case. Over the years, two generally strategic approaches to the problem have emerged: Prediction and risk minimization.
The former is based upon the thesis that, to a degree, movements in exchange rates are predictable and that returns can be increased by taking advantage of this element of predictability. The latter derives from the belief that the risks of running exposed positions in currencies are both large and uncontrollable and, therefore, every attempt should be made to minimize or eliminate risk, regardless of the cost to the rate of return. A strategy in which prediction plays a major part is essentially one that attempts, simultaneously to maximize return and minimize risk.
At the limit, if it is known with complete certainty that a given currency is going to yield the highest return in a subsequent period, then, by investing in it, the maximum return can be earned with no risk since the outcome will be exactly that which is expected. This utopian consequence of the ability to predict explains why so much effort is put into developing models to forecast exchange rates.
The rewards from making correct predictions are so great that forecasters who have a proven ability to predict exchange rates conceal this ability from the world at large.
When such methods exist, they are kept secret as they would generate large rewards for their users and ultimately, the investors.
The second general approach to currency selection is risk minimization. It is based upon the principle that no reward is large enough to compensate for the risks of exposure to shifts in exchange rates.
Therefore, the diversification of currencies in the portfolio is matched to the actual or implied liabilities with which the assets are associated. The portfolio is thus insulated from the effects of exchange rate fluctuation.
In essence, this strategy has become the widespread response to the volatility and unpredictability of foreign exchange markets.
Unfortunately, there are serious practical difficulties expressed with this strategy. The obvious conclusion, therefore, is that the two most widely used approaches, i.e. prediction and risk minimization, have failings at a theoretical and a practical level which render them insufficient to the task of selecting currencies in a multicurrency portfolio.
Consequently, some other approach has to be developed. There must be some logically determined framework within which the selection process takes place if the portfolio is to be managed in the true sense of the word.
Returning to some basic considerations, since by combining some fairly fundamental notions about asset management with some moderately self-evident observations about the way currencies behave, it is possible to arrive at an approach to choosing currencies which is defensible from the point of view of logic, and practical in the world of day-to-day portfolio management.
Conventionally expressed, the objective of all kinds of asset management is the maximization of return, subject to some constraint on risk. While this objective is not always made explicit, currency management requires the establishment of a balance between risk and return.
Nevertheless, the first step must be to measure them ‑ if only on a historical basis. Historically, there are some distinct differences in the way currencies behave suggesting a correlation between risk and return providing a semblance of predictability about them.
Another characteristic of exchange rates, which has an element of predictability is that, when they fluctuate, some do so by more than others. In other words, some have a history of greater volatility than others. Once again, the reason for this characteristic seems to lie in the nature of the economic linkages between countries.
From the above, one can observe certain relevant characteristics of the way in which currencies behave which have proven fundamental causes and reasonable element of predictability about them.
The next stage is to combine these elements with some basic ideas of portfolio management to arrive at an approach for selecting currencies. Whichever strategy the portfolio manager ultimately adopts, he will be viewing his choice of currencies in the context of its effect on his portfolio and not as a collection of wholly unrelated decisions governed by judgments about each currency separately.
(Habib F. Faris is vice president at Clariden Bank, London.)